Healthy diapering with passive RFIDs for diaper wetness sensing and urine pH identification

Abstract

In this paper, we present RFDiaper, a commodity passive RFID based healthy diapering system, which can sense the diaper wetness (i.e., wet/dry) and identify pH value of urine absorbed by the diaper. To do so, we leverage the coupling effect between the urine absorbed by the diaper and RFID tag, thereby the phase and amplitude variation can indicate urine pH and diaper wetness. However, rich scattering and dynamic environment exhibit a great challenge for accurate diaper wetness sensing and urine pH identification. Therefore, we propose a twin-tag based dynamic environment mitigation approach for robust and healthy diapering. Specifically, by extracting the differential amplitude and phase from the co-located sensing tag and reference tag (i.e., twin-tag) attached on the diaper, the multipath effect and the other dynamic factors (e.g., diaper wearer's body, tag's orientation and temperature, etc.) can be mitigated. Then, we detect the diaper wetness and estimate the urine pH based on differential amplitude and phase. We have implemented RFDiaper's design and evaluated its effectiveness with the experiments using commercial off-the-shelf (COTS) RFID tags attached on the diaper worn by the doll and the human subjects. RFDiaper can achieve the median accuracy of around 96% for diaper wetness sensing and urine pH estimation error of around 0.23 in dynamic environment. CCS CONCEPTS• Computer systems organization → Sensor networks; • Humancentered computing → Ubiquitous and mobile computing systems and tools; Human computer interaction (HCI).